Optimize development builds by default

.github/workflows/ci.yaml

@@ -19,6 +19,8 @@ defaults:
 env:
   CARGO_INCREMENTAL: 0
   CARGO_NET_RETRY: 10
+  CARGO_PROFILE_DEV_LTO: "false"
+  CARGO_PROFILE_DEV_OPT_LEVEL: 0
   CARGO_TERM_COLOR: always
   RUSTUP_MAX_RETRIES: 10
   PACKAGE_NAME: ruff

.github/workflows/daily_fuzz.yaml

@@ -22,6 +22,8 @@ env:
   RUSTUP_MAX_RETRIES: 10
   # Line-tables-only debug info: faster builds, backtraces still work.
   CARGO_PROFILE_DEV_DEBUG: line-tables-only
+  CARGO_PROFILE_DEV_LTO: "false"
+  CARGO_PROFILE_DEV_OPT_LEVEL: 0
   PACKAGE_NAME: ruff
   FORCE_COLOR: 1
 

.github/workflows/publish-docs.yml

@@ -20,6 +20,10 @@ on:
 permissions:
   contents: read
 
+env:
+  CARGO_PROFILE_DEV_LTO: "false"
+  CARGO_PROFILE_DEV_OPT_LEVEL: 0
+
 jobs:
   mkdocs:
     environment:

.github/workflows/publish-playground.yml

@@ -15,6 +15,8 @@ on:
 env:
   CARGO_INCREMENTAL: 0
   CARGO_NET_RETRY: 10
+  CARGO_PROFILE_DEV_LTO: "false"
+  CARGO_PROFILE_DEV_OPT_LEVEL: 0
   CARGO_TERM_COLOR: always
   RUSTUP_MAX_RETRIES: 10
 

.github/workflows/publish-ty-playground.yml

@@ -21,6 +21,8 @@ concurrency:
 env:
   CARGO_INCREMENTAL: 0
   CARGO_NET_RETRY: 10
+  CARGO_PROFILE_DEV_LTO: "false"
+  CARGO_PROFILE_DEV_OPT_LEVEL: 0
   CARGO_TERM_COLOR: always
   RUSTUP_MAX_RETRIES: 10
 

.github/workflows/sync_typeshed.yaml

@@ -44,6 +44,8 @@ env:
 
   CARGO_NET_RETRY: 10
   CARGO_PROFILE_DEV_DEBUG: line-tables-only
+  CARGO_PROFILE_DEV_LTO: "false"
+  CARGO_PROFILE_DEV_OPT_LEVEL: 0
   CARGO_TERM_COLOR: always
   NEXTEST_PROFILE: "ci"
   RUSTUP_MAX_RETRIES: 10

.github/workflows/typing_conformance.yaml

@@ -36,6 +36,8 @@ env:
   RUST_BACKTRACE: 1
   # Line-tables-only debug info: faster builds, backtraces still work.
   CARGO_PROFILE_DEV_DEBUG: line-tables-only
+  CARGO_PROFILE_DEV_LTO: "false"
+  CARGO_PROFILE_DEV_OPT_LEVEL: 0
   CONFORMANCE_SUITE_COMMIT: f4f2952f3ac94d7af819c5c71b60a50a100370e0
   PYTHON_VERSION: 3.12
 

AGENTS.md

@@ -21,25 +21,25 @@ instructions to PR authors or flag unrelated pre-existing issues.
 Run all tests (using `nextest` for faster execution and setting `INSTA_FORCE_PASS=1 INSTA_UPDATE=always MDTEST_UPDATE_SNAPSHOTS=1` to ensure all snapshots are updated):
 
 ```sh
-CARGO_PROFILE_DEV_OPT_LEVEL=1 CARGO_PROFILE_DEV_LTO=off INSTA_FORCE_PASS=1 INSTA_UPDATE=always CARGO_PROFILE_DEV_DEBUG="line-tables-only" MDTEST_UPDATE_SNAPSHOTS=1 cargo nextest run
+INSTA_FORCE_PASS=1 INSTA_UPDATE=always MDTEST_UPDATE_SNAPSHOTS=1 cargo nextest run
 ```
 
 Run tests for a specific crate:
 
 ```sh
-CARGO_PROFILE_DEV_OPT_LEVEL=1 CARGO_PROFILE_DEV_LTO=off INSTA_FORCE_PASS=1 INSTA_UPDATE=always CARGO_PROFILE_DEV_DEBUG="line-tables-only" MDTEST_UPDATE_SNAPSHOTS=1 cargo nextest run -p ty_python_semantic
+INSTA_FORCE_PASS=1 INSTA_UPDATE=always MDTEST_UPDATE_SNAPSHOTS=1 cargo nextest run -p ty_python_semantic
 ```
 
 Run a single mdtest file. The path to the mdtest file should be relative to the `crates/ty_python_semantic/resources/mdtest` folder. Include `--test mdtest` to avoid building unrelated test binaries:
 
 ```sh
-CARGO_PROFILE_DEV_OPT_LEVEL=1 CARGO_PROFILE_DEV_LTO=off INSTA_FORCE_PASS=1 INSTA_UPDATE=always CARGO_PROFILE_DEV_DEBUG="line-tables-only" MDTEST_UPDATE_SNAPSHOTS=1 cargo nextest run -p ty_python_semantic --test mdtest -- mdtest::<path/to/mdtest_file.md>
+INSTA_FORCE_PASS=1 INSTA_UPDATE=always MDTEST_UPDATE_SNAPSHOTS=1 cargo nextest run -p ty_python_semantic --test mdtest -- mdtest::<path/to/mdtest_file.md>
 ```
 
 To run a specific mdtest within a file, use a substring of the Markdown header text as `MDTEST_TEST_FILTER`. Only use this if it's necessary to isolate a single test case:
 
 ```sh
-MDTEST_TEST_FILTER="<filter>" CARGO_PROFILE_DEV_OPT_LEVEL=1 CARGO_PROFILE_DEV_LTO=off INSTA_FORCE_PASS=1 INSTA_UPDATE=always CARGO_PROFILE_DEV_DEBUG="line-tables-only" MDTEST_UPDATE_SNAPSHOTS=1 cargo nextest run -p ty_python_semantic --test mdtest -- mdtest::<path/to/mdtest_file.md>
+MDTEST_TEST_FILTER="<filter>" INSTA_FORCE_PASS=1 INSTA_UPDATE=always MDTEST_UPDATE_SNAPSHOTS=1 cargo nextest run -p ty_python_semantic --test mdtest -- mdtest::<path/to/mdtest_file.md>
 ```
 
 ### Fallback without nextest
@@ -48,16 +48,16 @@ If `cargo nextest` is not available, use `cargo test` with the same environment
 
 ```sh
 # Run all tests.
-CARGO_PROFILE_DEV_OPT_LEVEL=1 CARGO_PROFILE_DEV_LTO=off INSTA_FORCE_PASS=1 INSTA_UPDATE=always CARGO_PROFILE_DEV_DEBUG="line-tables-only" MDTEST_UPDATE_SNAPSHOTS=1 cargo test
+INSTA_FORCE_PASS=1 INSTA_UPDATE=always MDTEST_UPDATE_SNAPSHOTS=1 cargo test
 
 # Run tests for a specific crate.
-CARGO_PROFILE_DEV_OPT_LEVEL=1 CARGO_PROFILE_DEV_LTO=off INSTA_FORCE_PASS=1 INSTA_UPDATE=always CARGO_PROFILE_DEV_DEBUG="line-tables-only" MDTEST_UPDATE_SNAPSHOTS=1 cargo test -p ty_python_semantic
+INSTA_FORCE_PASS=1 INSTA_UPDATE=always MDTEST_UPDATE_SNAPSHOTS=1 cargo test -p ty_python_semantic
 
 # Run a single mdtest file.
-CARGO_PROFILE_DEV_OPT_LEVEL=1 CARGO_PROFILE_DEV_LTO=off INSTA_FORCE_PASS=1 INSTA_UPDATE=always CARGO_PROFILE_DEV_DEBUG="line-tables-only" MDTEST_UPDATE_SNAPSHOTS=1 cargo test -p ty_python_semantic --test mdtest -- <path/to/mdtest_file.md>
+INSTA_FORCE_PASS=1 INSTA_UPDATE=always MDTEST_UPDATE_SNAPSHOTS=1 cargo test -p ty_python_semantic --test mdtest -- <path/to/mdtest_file.md>
 
 # Run a specific mdtest within a file.
-MDTEST_TEST_FILTER="<filter>" CARGO_PROFILE_DEV_OPT_LEVEL=1 CARGO_PROFILE_DEV_LTO=off INSTA_FORCE_PASS=1 INSTA_UPDATE=always CARGO_PROFILE_DEV_DEBUG="line-tables-only" MDTEST_UPDATE_SNAPSHOTS=1 cargo test -p ty_python_semantic --test mdtest -- <path/to/mdtest_file.md>
+MDTEST_TEST_FILTER="<filter>" INSTA_FORCE_PASS=1 INSTA_UPDATE=always MDTEST_UPDATE_SNAPSHOTS=1 cargo test -p ty_python_semantic --test mdtest -- <path/to/mdtest_file.md>
 ```
 
 ### Snapshot updates
@@ -80,7 +80,7 @@ Never edit snapshot files or inline snapshot bodies manually. Regenerate them by
 ## Running Clippy
 
 ```sh
-CARGO_PROFILE_DEV_OPT_LEVEL=1 CARGO_PROFILE_DEV_LTO=off CARGO_PROFILE_DEV_DEBUG="line-tables-only" cargo clippy --workspace --all-targets --all-features -- -D warnings
+cargo clippy --workspace --all-targets --all-features -- -D warnings
 ```
 
 ## Running Debug Builds
@@ -90,13 +90,13 @@ Use debug builds (not `--release`) when developing, as release builds lack debug
 Run Ruff:
 
 ```sh
-CARGO_PROFILE_DEV_OPT_LEVEL=1 CARGO_PROFILE_DEV_LTO=off CARGO_PROFILE_DEV_DEBUG="line-tables-only" cargo run --bin ruff -- check path/to/file.py
+cargo run --bin ruff -- check path/to/file.py
 ```
 
 Run ty:
 
 ```sh
-CARGO_PROFILE_DEV_OPT_LEVEL=1 CARGO_PROFILE_DEV_LTO=off CARGO_PROFILE_DEV_DEBUG="line-tables-only" cargo run --bin ty -- check path/to/file.py
+cargo run --bin ty -- check path/to/file.py
 ```
 
 ## Working on ty
@@ -162,7 +162,7 @@ Parts of `.github/workflows/release.yml` are generated by cargo-dist from `dist-
 - Prefer let chains (`if let` combined with `&&`) and let guards (`PAT if let ... =>`) over nested `if let` statements to reduce indentation and improve readability. At the end of a task, always check your work to see if you missed opportunities to use `let` chains or `let` guards.
 - If you _have_ to suppress a Clippy lint, prefer to use `#[expect()]` over `[allow()]`, where possible. But if a lint is complaining about unused/dead code, it's usually best to just delete the unused code.
 - Don't use comments to narrate code, but do use them to explain invariants and why something unusual was done a particular way. Make sure that a comment will make sense to somebody who's reading the code for the first time. Prefer plain language, avoid jargon, and don't be afraid to be more verbose if it's necessary to explain something well. Giving examples of the kind of Python code we're trying to model at this particular point in Ruff or ty can often be very helpful for future readers of the code.
-- Run `CARGO_PROFILE_DEV_OPT_LEVEL=1 CARGO_PROFILE_DEV_LTO=off CARGO_PROFILE_DEV_DEBUG="line-tables-only" cargo dev generate-all` after changing configuration options, CLI arguments, lint rules, or environment variable definitions, as these changes require regeneration of schemas, docs, and CLI references.
+- Run `cargo dev generate-all` after changing configuration options, CLI arguments, lint rules, or environment variable definitions, as these changes require regeneration of schemas, docs, and CLI references.
 - Don't prefix tests with `test_`.
 - Don't separate struct definitions from their `impl` blocks unless the `impl` is deliberately placed in a separate file, as for large structs.
 - Avoid running `uv run` for any scripts from the repository root unless you use `--no-project`, `--script` or similar. Using `uv run` from the Ruff repo root without these flags will build Ruff from source, which is very slow and usually unnecessary.

Cargo.toml

@@ -300,6 +300,12 @@ inherits = "release"
 opt-level = "z"
 codegen-units = 1
 
+[profile.dev]
+opt-level = 1
+debug = "line-tables-only"
+# Avoid the local ThinLTO that Cargo enables at nonzero optimization levels.
+lto = "off"
+
 [profile.dev.package.insta]
 opt-level = 3